On the feasibility of generalized Interference Alignment with Partial Interference Cancelation

Abstract

We study the feasibility condition for a new IA strategy which is referred to as “Partial Interference Cancelation-based Interference Alignment” (PIC-IA). Unlike the conventional interference alignment (IA) strategy, PIC-IA aims to remove the most significant interference signals, and what we concern about the scheme is how many interference signals can be aligned and canceled for a system. And in the paper, we show that the feasibility is no longer limited by the number of users K. Two typical scenarios are considered in this paper, namely, a single-stream and a multi-stream system. And the graph theory is introduced to investigate the feasibility conditions under both scenarios. Compared with other existing works about the feasibility, our paper considers a more general scenario in which different users can transmit different numbers of data streams and the maximum number of interference streams to be removed for each user can be different. Thus the results are more meaningful to practical systems.